Micro, small or large: how the site questions change

Three useful starting points.

Every number above is an Atlas planning input or a calculation from those inputs. The site-area assumptions are screening prompts, not layout standards. Connectivity targets are requested services, not measured carrier availability. Two requested routes are not two verified independent routes.

MW means megawatts; one MW equals 1,000 kW. A micro project under one MW can still have substantial service, space, noise, fire-protection and reliability requirements. Size alone does not establish a simple connection.

Micro: prove the concept and protect the next step.

A 0.5 MW scenario is a useful starting point for pilot, edge or proof-of-concept planning. The first questions are often specific: what service already exists, what changes are necessary, how much room is available for supporting plant, and does the site allow the intended operation?

Use the whole property boundary even for a compact installation. Existing buildings, easements, equipment clearances, access and construction staging can matter more than gross acreage. Test the peak heat-rejection design rather than assuming a low annual water estimate means cooling is resolved.

Consider the next phase now. If the pilot is intended to become a larger operation, record ultimate demand and land needs in the brief. A favourable answer for the first module should not be carried forward as evidence for a larger build-out. Cryptocurrency mining has separate restrictions. The Province describes a permanent ban affecting certain new BC Hydro cryptocurrency operations; the micro preset does not establish an exemption. Source: B.C. cryptocurrency mining policy.

Small: design the service around the workload.

The 5 MW profile gives neocloud and regional-facility teams a starting point for a repeatable operating platform. The business model does not determine the electrical design: rack density, utilization, cooling, network traffic and resilience objectives still drive the brief.

At this scale, ask how the utility service and carrier connections will be delivered, not only whether infrastructure is mapped nearby. For the default two-route target, request physical route drawings and shared-dependency information. Different carrier names do not, by themselves, demonstrate separate ducts, entrances or upstream paths.

A dry-cooling scenario assumes no evaporative makeup in Atlas’s water sensitivity. It still leaves domestic demand, fire-flow requirements, any other process water and the design’s performance during heat extremes to be checked. If the cooling concept changes, rerun the water and electrical scenarios together.

Large: establish the allocation and campus strategy early.

The 25 MW profile is an initial campus-scale enquiry. It is not a claim that every hyperscale project has this capacity, or that ten hectares is sufficient for the final layout. Expansion, utility plant, setbacks, access and phasing can substantially change the land requirement.

For covered BC Hydro data-centre requests of at least 10 MW, the current competitive allocation process is a material early question. Confirm the applicable route, existing agreements and intended phases before relying on a delivery programme. Source: B.C. industrial electricity allocation framework.

The default hybrid cooling scenario intentionally introduces a water sensitivity. It should lead to questions about seasonal reliability, peak-day flow, treatment and discharge—not a conclusion that the nearest stream or water main can supply the campus. Read the water and fibre enquiry guide.

Compare like with like, then test the alternatives.

Run one consistent profile across the shortlist first. Next, change the operating scenario at the most promising sites and inspect what moves: energy, cooling water, connectivity targets, footprint needs and unresolved utility requirements.

Keep both sets of results. A small project that fits one development brief does not automatically become a suitable large campus. Atlas separates assumptions from observed public evidence so those changes remain explainable in the map and report.

For a worked example of how the same selected footprint produces different questions across the three profiles, see the public sample report. For the acquisition workflow, use the site-selection checklist.